I think we have to find a way to deal with this issue of names. Let us try H. magnifica. It originated in a population from the Frehse Reserve SE Riversdale and it is truly difficult to circumscribe all those individual variants.  The name has also been attached to a number of other plants and populations from various places.  The name magnifica is not clearly assignable beyond individuals that can be said to resemble the type (an illustration) and there are individual plants in the Frehse population that do not accord with either picture or description.  There are individual plants and populations going all the way to near Caledon that confound the name still further. In my opinion the Frehse reserve plants belong to a single system that I consider to be H. mirabilis (and I am not so sure that it is not bigger still).  Do we just drop the name ‘magnifica’ and use locality?   So is it better to say H. mirabilis magnifica (Frehse Reserve) and H. mirabilis magnifica (3km S Riversdale) or H. mirabilis magnifica (Windsor) for the variants that occur at each of those places?  Also H. mirabilis jakubii (Goukou) that I think is a connection between H. mirabilis magnifica  and H. mirabilis paradoxa (Vermaaklikheid and on to Infanta).  We could use this system and also the Breuer and Hayashi names attached to other mirabilis populations in the Riversdale area.  The disadvantage of place names is that they convey nothing to people unfamiliar with local geography and to them there may be no difference.  On the other hand that the formal Latin names may be restricted by the accompanying description and illustration, and not convey the variations that occur in the various populations.